Recent conceptual frameworks for understanding violence refer to the importance of community social relations or social capital. Pioneering research in the USA by Sampson et at (1997) has demonstrated the association between social capital and violent crime, which is independent of poverty levels. Intervention programs are now beginning to act upon this knowledge by attempting to strengthen social capital. However, little is known about how to effectively strengthen social capital and its effects on violence, or the relationship between social capital and mental health. The municipal health department of Cali (Colombia) is attempting to reduce the high levels of violence in the city by strengthening social capital. They are working together with a non-governmental organization, FUNDAPS, and CISALVA (the violence research group of Universidad del Valle) to promote reduced youth violence by working with youth organizations in three areas of the city. FUNDAPS is also working with health institutions and municipal departments (e.g. health, education, social welfare) to enable them to respond to youth needs. The proposed project aims to reduce youth violence through strengthening social capital. It involves measuring levels of social capital, violence and mental health in a baseline study and follow up after two years of intervention. Three intervention communities and three control communities are examined. The measure of social capital incorporates the latest conceptual developments in the field by explicitly measuring both cognitive and structural components of social capital. The tool for measuring perception and prevalence of violence has already been used with low-income communities in Cali by CISALVA. The measure of mental health will be the WHO recommended Self-Reporting Questionnaire, 20 items, which has high validity and reliability in such settings. Household surveys will be used to interview 100 youths (aged 15-25 years) in each of the six communities. Focus group discussions with youths of different ages (aged 10-25 years) and socio-economic status will be used to supplement household survey measures. Key informant interviews with youth organization leaders, municipal staff and staff of numerous government sectors will also be undertaken. The project enables cross-fertilization of research on youth violence between US, European and institutions in the South, which has been limited to date.